handle Unicode encoding in Console.Beep #105094
Conversation
|
Tagging subscribers to this area: @dotnet/area-system-console |
| bool isUnicode = Console.OutputEncoding.CodePage == UnicodeCodePage; | ||
|
|
||
| // Windows doesn't use terminfo, so the codepoint is hardcoded. | ||
| ReadOnlySpan<byte> bell = isUnicode ? stackalloc byte[2] { 7, 0 } : stackalloc byte[1] { 7 }; |
There was a problem hiding this comment.
I assume we don't need to worry about endianness here?
| @@ -679,8 +679,13 @@ public static void Beep() | |||
| { | |||
| if (!Console.IsOutputRedirected) | |||
| { | |||
| ReadOnlySpan<byte> bell = "\u0007"u8; // Windows doesn't use terminfo, so the codepoint is hardcoded. | |||
| int errorCode = WindowsConsoleStream.WriteFileNative(OutputHandle, bell, useFileAPIs: Console.OutputEncoding.CodePage != UnicodeCodePage); | |||
| bool isUnicode = Console.OutputEncoding.CodePage == UnicodeCodePage; | |||
There was a problem hiding this comment.
Are there other encodings that are not UnicodeCodePage and that may end up the bell characters into more than one byte? Would it make more sense to just encode the bell character using the OutputEncoding instead of trying to guess what the encoding may do here?
There was a problem hiding this comment.
You are right. Moreover, this method should never be on a hot path so we don't really gain anything by hardcoding the encoded bytes. I've changed the implementation, PTAL
| int errorCode = WindowsConsoleStream.WriteFileNative(OutputHandle, bell, useFileAPIs: Console.OutputEncoding.CodePage != UnicodeCodePage); | ||
| if (errorCode == Interop.Errors.ERROR_SUCCESS) | ||
| Span<byte> bell = stackalloc byte[10]; | ||
| if (Console.OutputEncoding.TryGetBytes("\u0007", bell, out int bytesWritten)) |
There was a problem hiding this comment.
OutputEncoding can be a custom encoding provided by the user. Nothing guarantees that it is going to encode the bell character into less than 10 bytes. For example, the encoding can be used as a hook to introduce some ridiculous escape sequences.
I would keep this simple a just let the encoding to allocate the array.
…gh to encode a single character for custom encoding
https://github.com/dotnet/runtime/pull/104966/files#r1679690576